Research Article

Chitinase 1 regulates pulmonary fibrosis by modulatingTGF-β/SMAD7 pathway via TGFBRAP1 and FOXO3Chang-Min Lee1, Chuan-Hua He1, Jin Wook Park1, Jae Hyun Lee2, Suchita Kamle1, Bing Ma1, Bedia Akosman1,Roberto Cotez1, Emily Chen1, Yang Zhou1, Erica L Herzog3, Changwan Ryu2, Xueyan Peng2, Ivan O Rosas4, Sergio Poli4,Carol Feghali Bostwick5, Augustine M Choi6, Jack A Elias1,7, Chun Geun Lee1

TGF-β1 is a critical mediator of tissue fibrosis in health and diseasewhose effects are augmented by chitinase 1 (CHIT1). However, themechanisms that CHIT1 uses to regulate TGF-β1–mediated fibroticresponses have not been defined. Here, we demonstrate that CHIT1enhances TGF-β1–stimulated fibrotic cellular and tissue responsesand TGF-β1 signaling. Importantly, we also demonstrate that theseeffects are mediated by the ability of CHIT1 to inhibit TGF-β1 in-duction of its feedback inhibitor, SMAD7. CHIT1 also interacted withTGF-β receptor associated protein 1 (TGFBRAP1) and forkhead boxO3 (FOXO3) with TGFBRAP1 playing a critical role in CHIT1 en-hancement of TGF-β1 signaling and effector responses and FOXO3playing a critical role in TGF-β1 induction of SMAD7. These pathwayswere disease relevant because the levels of CHIT1 were increasedand inversely correlated with SMAD7 in tissues from patients withidiopathic pulmonary fibrosis or scleroderma-associated interstitiallung disease. These studies demonstrate that CHIT1 regulates TGF-β1/SMAD7 axis via TGFBRAP1 and FOXO3 and highlight the impor-tance of these pathways in the pathogenesis of pulmonary fibrosis.

DOI 10.26508/lsa.201900350 | Received 18 February 2019 | Revised 29 April2019 | Accepted 30 April 2019 | Published online 13 May 2019

Introduction

TGF-β1 is a pleiotropic regulator of essential biologic responses(Massague, 1998). Central to its biology are its well-documentedroles in wound healing and its importance as a regulator of in-flammation, and cell survival, differentiation, motility, proliferation,and adhesion (Bowen et al, 2013). TGF-β1 dysregulation has beenimplicated in the pathogenesis of a wide variety of diseases and is apotent regulator of scarring in human diseases characterized byfibrogenesis (Bowen et al, 2013). In many of these disorders, theoverexpression of TGF-β1 has been documented. In other diseases,such as scleroderma (SSc), exaggerated levels of TGF-β1 are not

seen, but TGF-β1 signatures are appreciated (Sargent et al, 2010). Inkeeping with its multifaceted effector functions and complex rolesin health and disease, it is now clear that the production andeffector responses of TGF-β1 are under precise control (Bowen et al,2013). The factors that regulate TGF-β1 effector response, however,have not been adequately defined.

The GH 18 gene family contains true chitinases that bind andcleave chitin. Chitinase 1 (CHIT1, chitotriosidase) is the most readilyappreciated true chitinase in mammals and man. It can be found inthe circulation and tissues of normal controls and patients withdisorders characterized by inflammation and remodeling such asGaucher’s disease, sarcoidosis, chronic obstructive pulmonarydisease, and interstitial pulmonary fibrosis (Cho et al, 2015).However, its roles in these disorders are poorly understood be-cause chitin (the only known substrate of chitinases) does not existin mammals. Previous studies from our laboratory demonstratedthat the levels of CHIT1 activity are increased in patients with SScwith interstitial lung disease (SSc-ILD) where they correlate withdisease severity (Lee et al, 2012). In vivo and in vitro investigationsalso demonstrated that CHIT1 interacted with TGF-β1 to augmentTGF-β1 receptor–induced SMAD and MAPK and AKT signaling (Leeet al, 2012). These studies demonstrate that CHIT1 is a biomarker forTGF-β1 in SSc-ILD where it augments TGF-β1 signaling (Lee et al,2012). However, the mechanism(s) by which these fibrogenic effectsof CHIT1 are mediated have not been adequately defined.

The effects of TGF-β1 ligands are mediated by type 1 and type 2receptors (Huang & Chen, 2012). They activate and phosphorylatereceptor-regulated SMAD proteins (SMAD2 and 3), which form aheterometric complex with SMAD4. The complexes then translocateto the nucleus where it regulates gene expression and activatescanonical and noncanonical signaling pathways (Miyazawa &Miyazono, 2017). In keeping with the critical importance of theregulation of TGF-β1 effector responses, TGF-β1 signaling is regu-lated by a variety of inhibitory mechanisms (Miyazawa & Miyazono,2017). Inhibitory SMADs, in particular SMAD7, are induced by TGF-β1

1Molecular Microbiology and Immunology, BrownUniversity, Providence, RI, USA 2Department of Internal Medicine, Yonsei University College of Medicine, Seoul, South Korea3Section of Pulmonary, Critical Care, and Sleep Medicine, Department of Internal Medicine, Yale School of Medicine, New Haven, CT, USA 4Brigham and Women’s Hospital,Boston, MA, USA 5Department of Medicine, College of Medicine, Medical University of South Carolina, Charleston, SC, USA 6Weill Cornell Medicine Pulmonary and Critical CareMedicine, New York, NY, USA 7Division of Medicine and Biological Sciences, Brown University, Warren Alpert School of Medicine, Providence, RI, USA

Correspondence: chun_lee@brown.edu; jack_elias@brown.edu

© 2019 Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 1 of 14

on 24 April, 2020life-science-alliance.org Downloaded from http://doi.org/10.26508/lsa.201900350Published Online: 13 May, 2019 | Supp Info:

and feedback to control TGF-β1 signaling responses (Yan et al, 2009;Sureshbabu et al, 2016). In addition to TGF-β1, a variety of othercytokines and microRNA regulate SMAD7. However, a relationshipbetween CHIT1 and SMAD7 has not been defined.

To further understand the mechanisms that CHIT1 uses toregulate TGF-β1 response, in vivo and in vitro approaches wereused to define the roles of CHIT1 and themechanisms that it uses inpulmonary fibrosis. These studies demonstrate that CHIT1 aug-ments TGF-β1–stimulated fibroblast proliferation, myofibroblastdifferentiation, and canonical and noncanonical signaling. Theyalso demonstrate that these responses are mediated by the abilityof CHIT1 to inhibit SMAD7 via interaction with TGF-β receptor–associated protein 1 (TGFBRAP1) and FOXO3. Last, they highlight thedisease relevance of these findings by demonstrating the increasedlevels of CHIT1 and decrease in SMAD7 in lung tissues from patientswith idiopathic pulmonary fibrosis (IPF) or SSc-ILD.

Results

CHIT1 enhances TGF-β1–stimulated fibroblast responses in vitro

To determine if CHIT1 has significant effects on the TGF-β1–stimulated cellular responses that could contribute to fibrosis, we

stimulated normal human lung fibroblasts (NHLFs) with recombi-nant (r) CHIT1 alone, or in combination with TGF-β1. CHIT1 by itselfdid not stimulate fibroblast proliferation, but it did enhance TGF-β1–stimulated fibroblast proliferation as measured by WST-1 assay(Fig 1A). Similarly, CHIT1 did not alter fibroblast differentiation intomyofibroblasts as assessed by α-smooth muscle actin (α-SMA)accumulation, but it augmented the ability of TGF-β1 to enhanceα-SMA accumulation (Fig 1B). This suggests that CHIT1 enhancesTGF-β1–stimulatedmyofibroblast transformation. In accord with thesefindings, TGF-β1 stimulated the expression of vimentin, paxicillin,fibronectin, and other extracellular matrix (ECM) genes such aselastin and collagen, and these effects were significantly enhancedby CHIT1 (Fig 1C). Collectively, these results indicate that CHIT1enhances TGF-β1–stimulated fibroblast proliferation, myofibroblasttransformation, and ECM expression.

CHIT1 enhances and plays a critical role in TGF-β1–stimulatedfibrotic responses in vivo

Studies were next undertaken to determine if CHIT1 influenced TGF-β1–induced fibrotic responses in vivo. In these experiments, we firstcompared the regulation of ECM proteins in transgenic mice inwhich CHIT1 and TGF-β1 were overexpressed in the lungs, indi-vidually or in combination. We also compared the effects of

Figure 1. CHIT1 enhances TGF-β1-stimulatedfibroblast responses in vitro.NHLF were stimulated with recombinant (r) TGF-β1(rTGF-β, 10 ng/ml) and rCHIT1 (250 ng/ml) for 24 h thenthe effect of rTGF-β and rCHIT1 was evaluated. (A)Fibroblast cell proliferation assay by WST-1. (B)Immunohistochemistry (IHC) using anti–α-smoothmuscle actin; Con, controls with no rTGF-β or rCHIT1stimulation. (C) Expression of markers of myofibroblastsand extracellular protein accumulation assessed byreal-time quantitative reverse transcription PCR (qRT-PCR). β-actin was used as an internal control. The valuesin panels A and C represent mean ± SEM of triplicatedevaluations in a minimum of two separate experiments.Panel B is a representative IHC in a minimum of threeseparate experiments. *P < 0.05, **P < 0.01, t test. Bars inpanel B, 50 μM.

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 2 of 14

transgenic TGF-β1 in mice with wild-type or null mutant CHIT1alleles. In accord with our in vitro findings cited above, althoughtransgenic or null mutation of CHIT1 alone did not stimulate orsuppress, transgenic TGF-β1 was a significant stimulator and CHIT1and TGF-β1 interacted to further stimulate the expression of ECMproteins, including fibronectin, elastin, and type 1 collagen (Fig 2A).Importantly, the ability of TGF-β1 to stimulate ECM proteins was atleast partially dependent on CHIT1 because the stimulatory effectsof transgenic TGF-β1 were significantly decreased in mice with nullCHIT1 loci (Fig 2B and C). When viewed in combination, these studiesdemonstrate that CHIT1 is both necessary and sufficient for optimalTGF-β1–stimulated fibrotic ECM tissue response in the lungs.

CHIT1 augments TGF-β signaling in the lung

Studies were next undertaken to define the role(s) of CHIT1 in TGF-βsignaling. In these experiments, we compared the activation ofcanonical TGF-β1 signaling pathways in the lungs from transgenicmice in which CHIT1 and TGF-β1 were targeted to the lung, aloneand in combination. This approach was undertaken because pre-vious studies from our laboratory highlighted the exaggerated fi-brotic responses in the lungs from mice in which CHIT1 and TGF-β1were simultaneously expressed compared with mice in which eachwas expressed individually (Lee et al, 2012). As can be seen in Fig 2D,the present studies demonstrated that transgenic TGF-β1 increasedcanonical Smad2/3, MAPK/Erk, and Akt activation compared withwild-type control mice and activation of these signaling pathwayswas further enhanced in the co-expression of TGF-β1 and Chit1.These studies demonstrate that CHIT1 augments TGF-β1 signaling invivo.

CHIT1 inhibits the expression of inhibitory SMAD7

TGF-β1 is well known to induce SMAD7 which, in turn, feeds back tocontrol TGF-β1 signaling and effector responses (Zhao et al, 2000).The studies described above demonstrate that CHIT1 augmentsTGF-β1 signaling and fibrotic responses. This led us to hypothesizethat the effects of CHIT1 might be mediated via alterations in theability of TGF-β1 to induce SMAD7. To test this hypothesis, wecompared the expression of Samd7 in in the lungs from transgenicmice in which Chit1 and TGF-β1 were expressed, alone and incombination. In these experiments, transgenic TGF-β1 was a potentstimulator, and transgenic Chit1 did not significantly alter the ex-pression of Smad7 (Fig 2E). Importantly, when Chit1 and TGF-β1 weresimultaneously overexpressed, the ability of TGF-β1 to stimulateSamd7 was markedly ameliorated (Fig 2E). These studies demon-strate that CHIT1 inhibits the ability of TGF-β1 to induce inhibitorySMAD7. In so doing, they support the concept that the synergisticinteractions of CHIT1 and TGF-β1 are mediated, at least in part, bythe ability of CHIT1 to inhibit the SMAD7-mediated feedback in-hibition that normally controls TGF-β1 response.

CHIT1 interacts with TGFBRAP1 and FOXO3

Studies were next undertaken to further define the mechanismsthat CHIT1 uses to regulate TGF-β1 signaling and cell and tissueresponses. Because CHIT1 is a secreted protein that can be found in

both extracellular and intracellular compartments, we first un-dertook studies to define its potential binding partner(s). PotentialCHIT1-binding partners were first defined using yeast 2 hybrid (Y2H)screening assays with a lung cDNA library. This approach identifiedseveral CHIT1-binding proteins, including TGFBRAP1 and FOXO3(Table S1). TGFBRAP1 has been reported to be a chaperone for SMADsignaling (Wurthner et al, 2001) and FOXO3 is an transcription factorfor multiple genes that play critical roles in metabolism, cellularstress response, tissue remodeling, and disease progression (Nho& Hergert, 2014; Webb et al, 2016). In keeping with their potentialimportance, the results of the Y2H assays were further evaluatedusing co-immunoprecipitation (Co-IP) and co-localization assays.The Co-IP and immunoblot evaluations of cells transfected withCHIT1, TGFBRAP1, and FOXO3 demonstrated significant molecularinteractions between CHIT1 and TGFBRAP1 or FOXO3 (Fig 3A). Themolecular interaction between CHIT1 and TGFBRAP1 or FOXO3 wasfurther supported by the co-localization immunohistochemicalevaluations on the NHLF after stimulation with recombinant TGF-β(Fig 3B). We noted similar pattern of CHIT1 expression and co-localization with TGFBRAP1 or FOXO3 in epithelial cells and mac-rophages (data not shown). When viewed in combination, thesestudies demonstrate that CHIT1 interacts with TGFBRAP1 and FOXO3.

TGFBRAP1 plays a critical role in CHIT1 augmentation of TGF-β1signaling

To define the role(s) of TGFBRAP1 in CHIT1 regulation of TGF-βsignaling, gene-specific siRNA silencing was used. In these ex-periments, NHLF cells were stimulated with rTGF-β1 (10 ng/ml) andrCHIT1 (250 ng/ml), alone and in combination, for 1 h. Comparisonswere made of cells treated with TGFBRAP1 siRNA (which decreasedthe levels of TGFBRAP1 protein expression by >75%, Fig 4A) orscrambled controls. As noted above, the simultaneous treatment offibroblasts with recombinant CHIT1 with TGF-β1 enhanced theability of TGF-β1 to activate SMAD2 and MAPK/ERK and AKT. Im-portantly, these CHIT1-stimulated TGF-β1 responses were markedlydecreased in cells treated with TGFBRAP1 siRNA (Fig 4A). We alsonoted modest decreases in TGFβR1 expression with TGFBRAP1 si-lencing, suggesting that Chit1 regulates TGF-β receptor expressionvia TGFBRAP1, at least in part. Luciferease-tagged SMAD2 promotorassay further confirmed that silencing of TGFBRAP1 ameliorate theCHIT1-induced promoter activity of SMAD2 with TGF-β stimulation(Fig 4B). Collectively, these studies indicate that TGFBRAP1 plays anessential role in the ability of CHIT1 to augment canonical andnoncanonical TGF-β1 signaling and TGFβR expression.

TGF-β1 stimulates SMAD7 via FOXO3

The studies noted above demonstrate that CHIT1 interacts withFOXO3, and we identified a FOXO3-binding site in the promoterregion of SMAD7. This led us to investigate whether FOXO3 plays arole in TGF-β1 and or CHIT1’s regulation of SMAD7 expression. Inthese experiments, NHLF cells were treated with TGF-β1 and rCHIT1,alone or in combination, with and without siRNA silencing of FOXO3.The siRNA that were used in these experiments decreased thelevels of FOXO3 by >70% (Fig S1). As noted above, TGF-β1 treatmentincreased SMAD7 mRNA expression in a manner that was reduced

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 3 of 14

Figure 2. CHIT1 enhances and plays a critical role in TGF-β1–stimulated signaling and fibrotic responses in vivo.6–8-wk-old WT (−), TGF-β Tg (+), Chit1 Tg (+), and Chit1 Tg/TGF-β Tg mice were subjected to the evaluation. These mice were euthanized after 14 d of transgeneinduction of doxycycline containing drinking water (0.5 g/l). The expression levels of TGF-β1 and Chit1 in the bronchoalveolar lavage measured ELISA were about1.8 ± 0.2 ng/ml and 13 ng ± 2.3/ml for TGF-β Tg and Chit1 Tg, respectively. (A, B) Evaluation on the ECM-associated gene expression and collagen accumulation in the lungs ofWT and TGF-β Tg mice with overexpression of Chit1 (Chit1 Tg) (A) or null mutation of Chit1 (Chit1−/−) (B) using real-time qRT-PCR or Sircol Collagen Assay. β-actinwas used as an internal control. (C) Representative Mallory’s trichrome staining of the lungs from TGF-β Tg mice with null mutation of Chit1 (Chit−/−). (D, E) Western blot

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 4 of 14

by the simultaneous administration of CHIT1 (Fig 5A). In theseexperiments, the silencing of FOXO3 reduced TGF-β1–stimulatedSMAD7 expression to the levels comparable with those seen incontrol cells (Fig 5A). The role of FOXO3 in TGF-β1–stimulated SMAD7expression was further validated using SMAD7 promoter-luciferaseassays. In these experiments, HEK293 cells were transfected withSMAD7 promoter-luciferase constructs that contain a wild-type FOXO3consensus sequence (ProSMAD7WT) (GTAAAC, located −802 bp up-stream of SMAD7 translation initiation site) (Fig S2) and SMAD7 lu-ciferase activity was evaluated after treatment with recombinantTGF-β1 and CHIT1, alone or in combination. As shown in Fig 5B, TGF-β1 stimulated SMAD7 luciferase activity in a manner that was op-posed by co-administration of rCHIT1. As can also be seen in Fig 5B,this stimulation was SMAD7-dependent because SMAD7 luciferaseactivity was not induced by TGF-β1 in cells that had been trans-fected with a mutant SMAD7 promoter construct (ProSMAD7,(Foxo)),which did not contain a FOXO3 consensus sequence. When viewedin combination, these studies demonstrate that the transcriptionfactor FOXO3 plays a critical role in TGF-β1 stimulation of SMAD7expression.

CHIT1 reduces the levels of TGF-β–stimulated nuclearFOXO3/pFOXO3

Our studies demonstrate that TGF-β1 stimulates SMAD7 via FOXO3,whereas CHIT1 inhibits TGF-β1–stimulated SMAD7. This led us todetermine if the inhibitory effects of CHIT1 could be attributed to itsability to regulate FOXO3. To address this issue, we evaluated theexpression of Foxo3 and pFoxo3 in whole-cell lysates and thecellular nuclear and cytoplasmic fractions from the cells of lungsfrom WT mice and mice in which TGF-β1 and or Chit1 were targetedto the lungs. As shown in Fig 5C, transgenic TGF-β1 modestly in-creased the levels of Foxo3 and pFoxo3 in the whole-lung lysatesand their nuclear and cytoplasmic fractions. Although transgenicChit1 did not significantly alter the expression of Foxo3 and pFoxo3by itself, it did increase the levels of TGF-β1–stimulated Foxo3 orpFoxo3 in the whole-lung lysates (Fig 5C). This was mainly due to asignificant increase in cytoplasmic Foxo3 and pFoxo3 with de-creased nuclear fraction of these moieties (Fig 5C). These studiesdemonstrate that CHIT1 regulates the ability of TGF-β1 to inducethe nuclear transcription factor FOXO3 and suggests that CHIT1inhibits TGF-β1 induction of SMAD7 by decreasing nuclear FOXO3accumulation.

The levels of CHIT1 expression are increased in IPF where theycorrelate inversely with SMAD7

Studies were next undertaken to determine if our murine findingsare relevant to patients with IPF. Because CHIT1 augmented TGF-β1–induced fibrotic tissue responses by inhibiting SMAD7, wecharacterized the expression of CHIT1 and SMAD7 in tissues andcells from IPF patients and normal controls. As shown in Fig 6A, the

levels of mRNA encoding CHIT1 were significantly higher in thelungs from patients with IPF compared with normal controls. Inaccord with our murine studies, there was an inverse correlationbetween the levels of expression of SMAD7 and CHIT1 in the IPFpatients (Fig 6B). To further investigate the interactions betweenthese moieties, alveolar epithelial cells (alveolar type 1 [AT1] +alveolar type 2 cells [AT2]) were isolated from controls and IPFpatients, and the levels of mRNA encoding CHIT1 and SMAD7 werecharacterized. These evaluations revealed increased levels ofexpression of CHIT1, decreased levels of SMAD7 expression, and aninverse correlation between the expression of CHIT1 and SMAD7 incells from IPF patients (Fig 6 C–E). Interestingly, when AT1 and AT2cells were evaluated separately, CHIT1 and SMAD7 were similarlyregulated (Fig S3). As seen in our murine modeling systems, thesestudies demonstrate that, CHIT1 is up-regulated in patients withIPF where its expression correlates inversely with the expressionof SMAD7.

Decreased SMAD7 expression in the lungs from patients withSSc-ILD

Previous studies from our laboratory, demonstrated that the levelsof CHIT1 activity in the circulation of patients with SSc-ILD areincreased where they correlate with abnormal lung function (Leeet al, 2012). To determine if themurine and IPF findings noted aboveare also relevant to SSc-ILD, we also evaluated the expression ofSMAD7 in tissues from these patients and determined if the levels ofCHIT1 activity correlated with disease progression. As can be seen inFig 7A and B, the levels of expression of SMAD7 were decreasedin lungs from patients with SSc-ILD when compared with controls.In addition, the levels of circulating CHIT1 activity were increased inthe patients with progressive SSc-ILD compared with stable SSc-ILD(Fig 7C), and overall progression-free survival was decreased inpatients with increased levels of CHIT1 activity in a manner thatpersisted even after adjustment for the relevant covariates of age,gender, race, and smoking status (Fig 7D). These findings demon-strate that CHIT1 is induced, whereas Samd7 is inhibited in SSc-ILDand support the concept that the CHIT1/SMAD7 axis plays a criticalrole in the development and progression of pulmonary fibrosis inSSc-ILD.

Discussion

Although mammals do not have chitin or chitin synthase, sub-stantial levels of chitinases and chitinase-like proteins are noted inthe circulation as well as in local tissues (Lee et al, 2009, 2011; Cho etal, 2015). CHIT1, a major enzymatically active true chitinase, isproduced, stored, and secreted by macrophages and neutrophils(van Eijk et al, 2005) and plays important roles in innate immunehomeostasis (Elias et al, 2005). This can be appreciated in thepivotal roles it plays in host defenses against chitin-containing

analysis on the activation of TGF-β signaling (MAPK/Erk, Akt and, and Smad2) and Smad7 expression using lung lysates from WT and TGF-β and Chit1 Tg mice.β-actin was used as an internal control. The values in panels A and B represent the mean ± SEM of five mice each group. Panels D and E are representative Western blotsin a minimum of three separate experiments. *P < 0.05, **P < 0.01, ***P < 0.001, t test. Bars in panel C, 100 μm.

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 5 of 14

pathogens such as fungi, protozoa, and insects (Boot et al, 2001). Asa sensitive biomarker of macrophage activation, dysregulated ex-pression of CHIT1 in the circulation or local tissue has been re-ported in a variety of human diseases, including Gaucher’s disease,diabetes, sarcoidosis, inflammatory bowel disease, atherosclerosis,Alzheimer’s disease, and prostate cancer (Kanneganti et al, 2012;Elmonem et al, 2016). Recent studies from our laboratory and othersalso demonstrated that CHIT1 is dysregulated in lung diseasescharacterized by inflammation and remodeling such as bacterialinfection, asthma, chronic obstructive pulmonary disease, and pul-monary fibrosis (Lee et al, 2012; Cho et al, 2015; James et al, 2016; Honget al, 2018; Sharma et al, 2018). However, the specificmechanisms thatCHIT1 uses to contribute to the pathogenesis of specific disease havenot been fully understood. This is particularly true for CHIT1 inpulmonary fibrosis. Although we reported an important role of CHIT1in the pathogenesis of SSc-ILD, however, the exact mechanism thatCHIT1 uses to regulate pulmonary fibrosis has not been determinedand the relevance to other forms of ILD remains uncertain.

In this study, we demonstrated that CHIT1 synergistically en-hances TGF-β–stimulated fibroblast proliferation and myofibro-blast transformation, and CHIT1 is sufficient and necessary foroptimal TGF-β–stimulated ECM deposition in the lungs. We iden-tified TGFBRAP1 and FOXO3 as interacting partners of CHIT1 thatregulate TGF-β signaling and effector functions. Our studies further

demonstrated that CHIT1 enhances TGF-β–stimulated activation ofSMAD2/3, AKT, and MAPK/ERK pathways mainly through TGFBRAP1while inhibiting the expression of SMAD7, a prototype of inhibitorySMAD of TGF-β signaling, via interaction with FOXO3. In support ofthese findings, we also noted increased expression of CHIT1 in thelungs or circulation of the patients with IPF or SSc-ILD, and thelevels of CHIT1 were inversely correlated with SMAD7 expression inthe lungs of these patients. When viewed in combination, thesestudies demonstrated a significant role of CHIT1 in the patho-genesis of pulmonary fibrosis via regulation of canonical andnoncanonical TGF-β–stimulated signaling and SMAD7 expression.

IPF, a prototypic fibrotic disorder, is a progressive lung diseasecharacterized by epithelial damage, fibroproliferative matrix de-position, and parenchymal remodeling (Raghu, 1998; Selman et al,2001; Krein & Winston, 2002). TGF-β is believed to play an importantrole in this dysregulation because it is expressed in an exaggeratedfashion in IPF where, in contrast to controls, a sizable percentage isbiologically active (Khalil et al, 1996, 2001; Xu et al, 2003). The im-portant role that TGF-β may play in this disorder can be seen instudies that demonstrate that TGF-β is a critical mediator of pul-monary fibrosis after bleomycin injury (Nakao et al, 1999;Yehualaeshet et al, 2000) and that transgenic overexpression in thelung or high-dose adenoviral TGF-β1 transfer causes progressivepulmonary fibrosis in vivo (Sime et al, 1997; Kelly et al, 2003; Lee et al,

Figure 3. CHIT1 interacts with TGFBRAP1 and FOXO3.(A) Co-immunoprecipitation (IP) and immunoblot (IB)evaluations on the MLE12 epithelial cells transfectedwith CHIT1, TGFBRAP1, and FOXO3 expressing constructs.(B) Double-labeled fluorescent IHC to localizeexpression of CHIT1, TGFBRAP1, and FOXO3 in NHLF.Panels A and B are representative Western blots andIHC evaluations of three separate experiments,respectively. Bars in panel B, 50 μm.

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 6 of 14

2004) and IPF-like fibroblastic foci in in vitro explants (Xu et al,2003). It is also interesting to note that TGF-β requires other co-factors that activate or enhance TGF-β signaling in the develop-ment of persistent fibrosis in the lung or other organs (Mori et al,1999; Leask, 2006; Tatler & Jenkins, 2012). In this regard, presentstudies support a possibility that CHIT1 could be the one thatenhances the signaling and effector function of TGF-β in the

development of pulmonary fibrosis. In the previous studies, wedemonstrated that overexpression of IL-13 induced prominentpulmonary fibrosis with increased expression of both TGF-β andCHIT1 (Lee et al, 2012). We also demonstrated that null mutation ofCHIT1 without direct intervention of TGF-β significantly reducedIL-13–induced pulmonary fibrosis, suggesting a critical role of CHIT1in TGF-β effector function in tissue fibrosis (Lee et al, 2012). Col-lectively, these findings suggest that CHIT1 plays an essential role inthe pathogenesis of pulmonary fibrosis in which CHIT1 and TGF-βare dysregulated. In this regard, it is reasonable to speculate thatCHIT1 also could play an important role in the pathogenesis of otherorgan fibrosis including sarcoidosis and nonalcoholic steatohe-patitis that showed dysregulated levels of both CHIT1 and TGF-β(Peverill et al, 2014; Casanova et al, 2015).

Recently, Van Dyken et al (2017) reported that acidic mammalianchitinase (AMCase) plays a protective role in spontaneously de-veloped pulmonary fibrosis related to chitin polymer accumulationin the lung. AMCase and CHIT1 are two true chitinases in mammalsand have different cellular and tissue expression and physiologicfunction, although they share similar chitinolytic enzyme activitywith different reaction conditions. It has been shown that AMCase,but not CHIT1, is prominently expressed in themurine lung, whereasthe expression of Chit1 (both in endogenous and induced ex-pression) is predominant in human lungs (Boot et al, 2005). Becausethe physiologic role of AMCase in the lungs is more likely adapted tothe chitin-rich environment, it is reasonable to speculate thatdecreased chitinase activity of AMCase in murine lung results intochitin accumulation. However, in human lungs, there is no en-dogenous source of chitin, and CHIT1 enzyme activity was reportedto be increased with ageing (Kurt et al, 2007) and in the lungs of IPF(Bargagli et al, 2007). These studies suggest that enzymatic de-ficiency of AMCase may not fully explain human lung pathologieswithout simultaneous consideration of Chit1, the major true chi-tinase expressed in human lungs. In addition, recent studies alsosupport a direct pathogenic role of chitinase-like proteins such asCHI3L1 that have no chitinase enzyme activity with retained chitin-binding ability, in tissue remodeling and fibrosis (Zhou et al, 2014,2015). These studies suggest that enzyme activity of either AMCaseor CHIT1 might not be the sole responsible factor for determiningfinal outcome of tissue response associated with chitinase dys-regulation. When viewed in combination, present studies suggestthat the relative contribution and potential interaction of AMCaseand CHIT1 in the pathogenesis pulmonary fibrosis need to befurther determined in future studies.

In this study, we demonstrate that CHIT1 interacts with in-tracellular proteins TGFBRAP1 or FOXO3 and regulates major TGF-βsignaling and SMAD7 expression. As a chaperone for SMAD4,TGFBRAP1 has been shown to associate with inactive heteromericTGF-β and activin receptor complexes, mainly through the type 2receptor, and is released upon activation of signaling and recruitSMAD4 to the vicinity of the receptor complex to facilitate its in-teraction with receptor-regulated SMADs, such as SMAD2 (Wurthneret al, 2001). However, the specific role of TGFBRAP1 in the patho-genesis of fibrosis in the lungs or other organs has not beenstudied. This is the first study showing TGFBRAP1 interacts withCHIT1 and mediates synergistic effect of CHIT1 on TGF-β–stimulatedcanonical and noncanonical signaling. It is also interesting to note

Figure 4. TGFBRAP1 plays a critical role in CHIT1 augmentation of TGF-β1signaling and TGFβR expression.(A)Western blot evaluation on the NHLF cells stimulated by rTGF-β and/or rCHIT1with (+) and without (−) siRNA silencing of TGFBRAP1. Scrambled siRNA was used ascontrol for TGFBRAP1 specific siRNA. (B) Evaluation on the SMAD activities on thecells transfected SMAD2-luciferase construct after stimulated by rTGF-β (20 ng/ml) and/or rCHIT1 (100 or 200 ng/ml) with and without siRNA silencing ofTGFBRAP1. Panel A is a representative immunoblot of three separate experiments.Values in panel B represent the mean ± SEM of triplicated evaluations in aminimum of two separate experiments. *P < 0.05, **P < 0.01, one-way ANOVA.

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 7 of 14

that silencing of TGFBRAP1 itself on normal lung fibroblasts min-imally alters the TGF-β activation of SMAD2 or MAPK/ERK, but CHIT1effects on these TGF-β signaling activation were abrogated, sug-gesting that TGFBRAP1 is a fine tuner of TGF-β signaling and effectorfunction in the lung. It is also interesting to note that there ismodest decrease in the expression of TGFβR1 expression withsilencing of TGFBRAP1, suggesting that CHIT1 regulates TGF-βreceptor expression via TGFBRAP1, at least in part.

Because silencing of TGFBRAP1 did not alter CHIT1 inhibition ofTGF-β–stimulated SMAD7 expression, we sought other mecha-nism(s) responsible for CHIT1 regulation of SMAD7 expression. Our

studies demonstrated that FOXO3, a transcriptional factor that hasa binding consensus sequence in the promoter region of SMAD7,plays a critical role in TGF-β and CHIT1-regulated expression ofSMAD7. TGF-β itself induced the expression and phosphorylationof both cytoplasmic and nuclear FOXO3. Interestingly, CHIT1 in-creases the levels of pFOXO3 in the cytoplasmic compartment withdepletion pFOXO3 in nuclear fraction. These studies suggest thatCHIT1 plays a role in nuclear/cytoplasmic shuttling and cyto-plasmic retention of pFOXO3, which might limit the accessibility ofFOXO3 in the nuclear fraction, resulted into decreased tran-scriptional activity of FOXO3 for SMAD7 or other FOXO3 target

Figure 5. TGF-β stimulates SMAD7 expression viaFOXO3.(A) Real-time qRT-PCR evaluations on the NHLF cellsafter stimulation with rTGF-β (10 ng/ml) and rCHIT1 (250ng/ml) with (+) and without (−) siRNA silencing ofFOXO3. (B) HEK293 cells transfected with luciferase-tagged SMAD7 promoter construct with (ProSMAD7WT;WT) and without (ProSMAD7,(FOXO3); mutant form)FOXO3 consensus sequence were stimulated with rTGF-β and/or rCHIT1, and then luciferase activity wasmeasured. (C) Western blot evaluations on the pFOXO3and FOXO3 expression in the lungs of WT (−), TGF-β Tg(+), CHIT1 Tg (+), and CHIT1/TGF-β Tg mice. Whole-lunglysates were further separated into cytoplasmic andnuclear fractions and subjected to the immunoblotsusing anti-pFOXO3 and FOXO3 antibodies. Values inpanels A and B represent the mean ± SEM of triplicatedevaluations in a minimum of two separate experiments.Panel C is a representative immunoblot of threeseparate experiments. H3, histone H3, used as internalcontrol of nuclear protein. **P < 0.01, ***P < 0.001, one-way ANOVA. ns, not significant.

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 8 of 14

genes. Because our studies show that FOXO3 is essential for theexpression of SMAD7, these results provide a potential mechanismto explain the inverse relationship between the levels of CHIT1 andSMAD7 expression noted in the lungs of IPF patients. In the pre-vious studies, fibroblasts established from the patients with IPFshowed that pathologic alteration of FOXO3 with reduced ex-pression contribute to the progression of IPF by protecting fi-broblasts from apoptotic or autophagic cell death responses (Imet al, 2015; Nho et al, 2011, 2013). It has been shown that Akt phos-phorylation of FOXO3 resulted into nuclear cytoplasmic transport,and subsequent ubiquitination of this phosphorylated form ofFOXO3 after binding with 14-3-3 protein is the major pathway oflowering FOXO3 levels in the fibroblasts or other cells (Kato et al,2006; Tzivion et al, 2011; Im et al, 2015). A recent study also reporteda critical role of FOXO3 in fibrogenesis and reduced expression ofFOXO3 was implicated in the pathogenesis of IPF (Al-Tamari et al,2018). Thus, current data are consistent with these findings in thesense that CHIT1 could contribute to the progression of pulmonaryfibrosis via alteration of nuclear FOXO3 levels. However, themechanism of CHIT1 regulation of cytosolic retention of pFOXO3 isnot clear yet. Because increased levels of CHIT1 in fibrotic lungs of

bleomycin-challenged mice were noted mainly in the cytoplasmicand perinuclear area of the cells (Fig S4), it is intriguing tospeculate that physical binding between CHIT1 and pFOXO3 couldbe a reason for enhanced cytoplasmic retention of this moiety inthe lung. It is also possible that CHIT1 uses other mechanism ofpFOXO3 regulation such as interaction with 14-3-3 protein thatfacilitates the nuclear/cytosolic shuttling of pFOXO3 (Tzivion et al,2011). These possibilities remain to be determined in the futurestudies.

SMAD7, a major inhibitory SMAD of TGF-β signaling, blocks thefunction of SMAD2 and SMAD3 by inhibiting their binding to theactivated receptors and exerting its negative effects on TGF-β1/SMAD signaling (Kavsak et al, 2000; Wang et al, 2016). The specificantifibrotic regulatory role of SMAD7 in the pathogenesis of pul-monary fibrosis has been demonstrated in various conditions.SMAD7 is the target of microRNAs associated with myofibroblastdifferentiation and bleomycin-induced lung fibrosis (Liu et al, 2010;Wang et al, 2016) and mediates IL-7 suppression of pulmonary fi-brosis (Huang et al, 2002). In this regard, present studies provideCHIT1 as a novel negative regulator of SMAD7, which enhancesTGF-β signaling and effector function.

Figure 6. Increased expression levels of CHIT1 in the patients with IPF where they correlate inversely with SMAD7.(A) Real-time qRT-PCR evaluations on the expression levels of CHIT1 in whole-lung lysates from normal controls and IPF patients. (B) A scatter plot with regression analysisdemonstrating inverse relationship between the levels of CHIT1 and SMAD7 expression in the lungs of IPF patients (Spearman rank test, P = 0.04). (C, D) qRT-PCRevaluations on CHIT1 and SMAD7 expression in alveolar epithelial cells (alveolar type 1 epithelial cells [AT1] and alveolar type 2 epithelial cells [AT2]) isolated from the lungsof normal controls and IPF patients. (E) A scatter plot with regression analysis demonstrating inverse relationship between the levels of CHIT1 and SMAD7 in alveolarepithelial cells isolated from the lungs of IPF patients (P = 0.036, Spearman’s rank test).

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 9 of 14

It is interesting to note that dysregulated CHIT1 activity in theserum was significantly associated with progressive SSc-ILDcompared with stable cases and that the levels of CHIT1 activitywere correlated with survival rates of the patients. In the evaluationof end stage lungs from IPF patients, we also see prominent in-creases of CHIT1 expression with suppression of SMAD7 expression.These studies suggest that CHIT1/SMAD7 axis plays an importantrole in disease progression or exacerbation rather than initiator offibrotic tissue response, and potential use of CHIT1/SMAD7 activityand expression both in serum and lung or BAL as a biomarker topredict disease severity and prognosis of patients with IPF andSSc-ILD.

In conclusion, these studies demonstrated that CHIT1 interactswith TGFBRAP1 and FOXO3 and enhances TGF-β–stimulated acti-vation of canonical and noncanonical signaling and the expressionof inhibitory SMAD7. Thus, CHIT1 and its interacting partners ofTGFBRAP1 or FOXO3 could be reasonable therapeutic targets for theintervention of patients with pulmonary fibrosis in that TGF-β andCHIT1 are significantly dysregulated.

Materials and Methods

Mice

C57 BL/6 mice were purchased from the Jackson Laboratory andwere housed at Brown University animal facilities. TGF-β Tg, Chit1−/−,and Chit1 Tg mice on C57 BL/6 background were generated and

maintained in our laboratory as previously described (Lee et al,2004, 2012). Both TGF-β Tg and Chit1 Tg mice are lung-specific in-ducible transgenic mice, and transgene expression was induced bydrinking water containing doxycycline (0.5 g/l). The mice with nullmutation or overexpression of Chit1 did not show any apparentabnormal phenotypes and developmental and signaling issues. Allmurine procedures were approved by the Institutional Animal Careand Use Committees at Brown University.

Cell culture

NHLF cells and human embryonic kidney 293 cells (HEK293) wereobtained from Lonza and American Type Culture Collection (ATCC),respectively. NHLF cells were cultured in fibroblast growth medium(Lonza CC-2512B) at 37°C in 5% CO2. HEK293 cells were maintained inDMEM supplemented with 10% FBS and 100 U/ml penicillin and 100μg/ml streptomycin at 37°C in 5% CO2 incubator. NHLF cells wereused within 10 passages according to the manufacturer's recom-mendation. MLE12 murine-transformed lung epithelial cells werepurchased from ATCC and cultured andmaintained using DMEM-F12medium supplemented with 10% FBS as recommended.

siRNA silencing of TGFBRAP-1 and FOXO3

siRNAs specific for human TGFBRAP1 (SC-36720; Santa Cruz) andFOXO3 (SC-37887; Santa Cruz) were used for silencing the levels ofmRNA encoding TGFBRAP1 and FOXO3 as per the manufacturer’sinstructions. NHLF cells were seeded on six-well plates andtransfected the next day with TGFBRAP1, FOXO3, or control siRNAs.

Figure 7. Decreased SMAD7 expression in the lungsfrom patients with SSc-ILD.(A) IHC evaluation on the SMAD7 expression in the lungsections from normal control and patients with SSc-ILD.SMAD7 staining of SSc-ILD lungs (right) revealsdecreased number of Samd7-positive cells especially inthe stromal region as compared with normal control(left) that showed a number of parenchymal cells andalsomacrophages are strongly stained. (B)Quantitationof SMAD7-positive (+) cells in the lung sections fromcontrols and SSc-ILD per high-powered field. (C) SerumCHIT1 activity in the SSc-ILD patients with stable andprogressive lung function. “Progressive” was defined inthe patients with more than 10% of absolute drop offorced vital capacity (FVC). (D) Survival analysis basedon the levels of chitinase activity (below and abovemedian value of 53.54 nM/ml/h) in the patients withSSc-ILD. SSc-ILD subjects (n = 56) with values exceedingthe median value of 53.54 (dotted line) demonstratereduced progression-free survival as defined byabsolute loss of %FVC within 2 years of enrollment.Scale bars, 30 μm, **P < 0.01, ***P < 0.001, Spearman’srank test. n, number of individuals subjected to theevaluation.

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 10 of 14

The cells were collected at the indicated time points and weresubjected to real-time (RT)-PCR or Western blot evaluations.

Co-IP and IB analysis

To assess interactions among CHIT1, TGFBRAP-1, and FOXO3, MLE12murine lung epithelial cells were co-transfected with TGFBRAP1 andFOXO3 plasmids (pcDNA3.1). Lysates from these cells were sub-jected to immunoprecipitation using anti-hCHIT1 mouse mono-clonal antibody (1/500 dilution, AF3559, R&D system). The Catch andRelease V2.0 (Reversible Immunoprecipitation System; EMD Milli-pore) kit was used for Co-IP according to the manufacturer’s in-struction. The precipitates were then evaluated by immunoblottingwith antibodies against CHIT1, TGFBRAP1, or FOXO3.

Co-localization with double-label IHC

To localize the expression of CHIT1, TGFBRAP1, and FOXO3, double-label IHC was undertaken with a modification of procedures de-scribed previously by our laboratory (Lee et al, 2016). NHLF cells (1 ×106 cells/well) were cultured overnight in four-well chamber slides(Falcon) at 37°C in 5% CO2 incubator, then washed with PBS. Tounmask antigens, the slides were placed for 20 min at high tem-perature under high pressure in citrate buffer (10 mM sodiumcitrate, 0.05% Tween 20, pH 6). The slides were then blocked with anonserum protein-blocking reagent (DakoCytomation Inc) for 1 h atroom temperature and incubated with primary antibodies (anti-CHIT1 [1/50 dilution, AF-5325; R&D system], anti-TGFBRAP1 [1/50dilution, SC-13134; Santa Cruz Biotechnology], and anti-FOXO3 [1/50dilution, 12829S; Cell Signaling]) for 60min at room temperature in ahumid chamber. Substitution of the primary antibody with PBSserved as a negative control. The slides were then washed in PBST(0.01% Tween 20) and incubated with secondary antibodies con-jugated with horseradish peroxidase (Cell Signaling). Reactionproducts were developed using DAB containing 0.3% hydrogenperoxide up to 5 min. Cell nuclei were stained with mountingmedium with DAPI (50 μl, H-1200; Vector Laboratories). Fluores-cence was detected by Immunofluorescence microscopy.

Evaluation of TGF-β signaling

Canonical SMAD activity of TGF-β1 signaling pathway was assessedusing dual reporter assays (SABiosciences). A SMAD-responsivefirefly luciferase construct that constitutively expressed Renillaluciferase was transfected into NHLF cells. After stimulation withrecombinant CHIT1 and TGF-β1, alone or in combination, SMADactivation was assessed by measuring the dual luciferase activities.SMAD2 activation and MAPK/ERK, AKT activation were evaluated byWestern blot as previously described (Lee et al, 2004). The ex-pression of SMAD7 was evaluated by immunoblot assay using anti-SMAD7 polyclonal rabbit antibody (Santa Cruz Biotechnology).

Evaluation of FOXO3 and SMAD7 interaction

HEK293 cells were seeded at a density of 1 × 106 cells/well in six-welldishes and grown to 70% confluence. For each well, 1 μg of pEZX-PG04-SMAD7 promoter construct (GeneCopoeia, Inc.) or 1 μg of the

FOXO3-binding site point-mutated SMAD7 promoter were trans-fected into the cells with the Lipofectamine 3000 transfectionreagent. After stimulation with rCHIT1 and rTGF-β1, alone or incombination, SMAD7 expression was assessed by measuring theSecrete-Pair Dual Luminescence Assay kit (GeneCopoeia, Inc.) andwas normalized with secreted alkaline phosphatase activitywithin each sample.

Evaluation on the levels of cytosolic and nuclear pFOXO3 andFOXO3

Phosphorylation and cytoplasmic/nuclear shuttling are the majorpathways regulating transcriptional activity of FOXO3 (Nho & Hergert,2014). Hence, studies are undertaken to see whether CHIT1 is im-plicated in this process. For this evaluation, cytosolic and nuclearfraction of the whole-lung lysates were prepared using nuclearcytoplasmic extraction reagents (Thermo Fisher Scientific) accordingto the procedures provided by the company and subjected to im-munoblot evaluation using anti-pFoxO3 (Serine253) (Cell Signaling)and FoxO3 (D19A7) (Cell Signaling).

Profiling of IPF and control lung tissues

Fresh human lung tissue was procured from subjects with end-stage lung disease, undergoing transplant or donor lungs that werenot implanted at the time of transplant. The demographics andclinical characteristics of the groups can be seen in Table S2. Allprotocols were approved by the Partners Institutional ReviewBoard. For each subject, small pieces (<5 mm) of whole-lung tissuewere treated with RNAlater (QIAGEN) and snap-frozen in liquidnitrogen before storage at −80°C.

Isolation of alveolar epithelial cell type 1 and type 2

Fresh human lung explants were sliced and washed with coldsterile PBS. Visible airway structures, vessels, blood clots, andmucin were removed. Lung specimens were minced mechanicallyinto small pieces (<1 mm3) and then incubated for 45 min in 37°C inmodified medium containing DMEM/F12 with added digestionenzymes: Liberase thermolysin medium (Roche), elastase (EC-134;Elastin Products Company), and DNase I (Thermo Fisher Scientific).Digested tissue was filtered using a metal strainer. 10% FBS wereadded to the flow-through to stop the enzymatic reaction. Thepellet was resuspended in DMEM/F12 medium and filtered using a100-μm strainer. The cells were resuspended in freezing medium(10% FBS and 10% DMSO in DMEM/F12), aliquoted, and stored inliquid nitrogen. Thawed single-cell suspension was sorted usingFACS. Sorting gate for AEC type 1 and type 2 were designed asfollows: DAPI−/CD45−/EpCAM+/PDPN+ and DAPI−/CD45−/EpCAM+/PDPN−, respectively. Isolated cells from 10 subjects diagnosed withIPF and 10 controls were used. Patient demographics and clinicalinformation can be found in Table S3.

Cohort of SSc-ILD patients

These studies were performed with HIC approval at the Yale Schoolof Medicine. Samples from subjects with SSc-ILD according to

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 11 of 14

current EULAR/ACR criteria were used for these studies. As shown inTable S4, subjects with CHIT1 levels exceeding the median level of52.39 nM/ml/h weremore likely to be older and to have smoked butwere otherwisematched for gender, race, disease subtype, and lungfunction.

Measurement of CHIT1 activity

CHIT1 enzyme activity was assessed as described previously in ourlaboratory (Lee et al, 2012).

Localization of SMAD7 expression in SSc

IHC was undertaken to localize SMAD7 protein using polyclonalrabbit anti human SMAD7 obtained from Santa Cruz Biotechnologyas described in our laboratory (Lee et al, 2009). Lung explants frompatients with SSc-ILD and nonfibrotic controls were subjected tothe staining and SMAD7 expressing cells were counted in sixrandomly chosen high power fields. The values averaged and themean ± SEM calculated for SSc-ILD versus normal lung.

Statistical analysis

Normality of data was assessed using the D’Agostino–Pearsonomnibus normality test. Normally distributed data are expressed asthe means ± SEM and assessed for significance by a t test or ANOVA,as appropriate. Categorical variables were compared using theFisher’s exact test. Nonparametric correlation between two vari-ables was evaluated by Spearman’s rank test.

Supplementary Information

Supplementary Information is available at https://doi.org/10.26508/lsa.201900350.

Acknowledgements

This work was supported by National Institute of Health grants U01 HL10863(JA Elias), PO1 HL114501 (JA Elias, AM Choi), and R01 HL115813 (CG Lee) R01HL109233, HL125250 (EL Herzog) from National Heart, Lung, Blood Institute.

Author Contributions

CM Lee: conceptualization, resources, data curation, validation, andinvestigation.C-H He: resources and methodology.JW Park: data curation and investigation.JH Lee: conceptualization, formal analysis, and methodology.S Kamle: resources, data curation, investigation, and methodology.B Ma: data curation and methodology.B Akosman: data curation and methodology.R Cortez: data curation and methodology.E Chen: data curation and investigation.Y Zhou: resources, formal analysis, and methodology.

EL Herzog: resources, data curation, funding acquisition, andwriting—original draft.C Ryu: resources and data curation.X Peng: resources and data curation.IO Rosas: resources and data curation.S Poli: resources and data curation.CF Bostwick: resources.AM Choi: data curation and funding acquisition.JA Elias: conceptualization, resources, data curation, supervision,funding acquisition, and writing—original draft, review, and editing.CG Lee: conceptualization, resources, data curation, formalanalysis, supervision, funding acquisition, methodology, andwriting—original draft, review, and editing.

Conflict of Interest Statement

The authors declare that they have no conflict of interest.

References

Al-Tamari HM, Dabral S, Schmall A, Sarvari P, Ruppert C, Paik J, DePinho RA,Grimminger F, Eickelberg O, Guenther A, et al (2018) FoxO3 animportant player in fibrogenesis and therapeutic target for idiopathicpulmonary fibrosis. EMBO Mol Med 10: 276–293. doi:10.15252/emmm.201606261

Bargagli E, Margollicci M, Luddi A, Nikiforakis N, Perari MG, Grosso S, PerroneA, Rottoli P (2007) Chitotriosidase activity in patients with interstitiallung diseases. Respir Med 101: 2176–2181. doi:10.1016/j.rmed.2007.05.008

Boot RG, Blommaart EF, Swart E, Ghauharali-van der Vlugt K, Bijl N, Moe C,Place A, Aerts JM (2001) Identification of a novel acidic mammalianchitinase distinct from chitotriosidase. J Biol Chem 276: 6770–6778.doi:10.1074/jbc.m009886200

Boot RG, Bussink AP, Verhoek M, de Boer PA, Moorman AF, Aerts JM (2005)Marked differences in tissue-specific expression of chitinases inmouse and man. J Histochem Cytochem 53: 1283–1292. doi:10.1369/jhc.4a6547.2005

Bowen T, Jenkins RH, Fraser DJ (2013) MicroRNAs, transforming growth factorbeta-1, and tissue fibrosis. J Pathol 229: 274–285. doi:10.1002/path.4119

Casanova N, Zhou T, Knox KS, Garcia JGN (2015) Identifying novel biomarkersin sarcoidosis using genome-based approaches. Clin Chest Med 36:621–630. doi:10.1016/j.ccm.2015.08.005

Cho SJ, Weiden MD, Lee CG (2015) Chitotriosidase in the pathogenesis ofinflammation, interstitial lung diseases and COPD. Allergy AsthmaImmunol Res 7: 14–21. doi:10.4168/aair.2015.7.1.14

Elias JA, Homer RJ, Hamid Q, Lee CG (2005) Chitinases and chitinase-likeproteins in T(H)2 inflammation and asthma. J Allergy Clin Immunol 116:497–500. doi:10.1016/j.jaci.2005.06.028

Elmonem MA, van den Heuvel LP, Levtchenko EN (2016) Immunomodulatoryeffects of chitotriosidase enzyme. Enzyme Res 2016: 2682680.doi:10.1155/2016/2682680

Hong JY, Kim M, Sol IS, Kim KW, Lee CM, Elias JA, Sohn MH, Lee CG (2018)Chitotriosidase inhibits allergic asthmatic airways via regulation ofTGF-beta expression and Foxp3(+) Treg cells. Allergy 73: 1686–1699.doi:10.1111/all.13426

Huang F, Chen YG (2012) Regulation of TGF-beta receptor activity. Cell Biosci 2:9. doi:10.1186/2045-3701-2-9

Huang M, Sharma S, Zhu LX, Keane MP, Luo J, Zhang L, Burdick MD, Lin YQ,Dohadwala M, Gardner B, et al (2002) IL-7 inhibits fibroblast TGF-beta

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 12 of 14

production and signaling in pulmonary fibrosis. J Clin Invest 109:931–937. doi:10.1172/jci14685

Im J, Hergert P, Nho RS (2015) Reduced FoxO3a expression causes lowautophagy in idiopathic pulmonary fibrosis fibroblasts on collagenmatrices. Am J Physiol Lung Cell Mol Physiol 309: L552–L561.doi:10.1152/ajplung.00079.2015

James AJ, Reinius LE, Verhoek M, Gomes A, Kupczyk M, Hammar U, Ono J, OhtaS, Izuhara K, Bel E, et al (2016) Increased YKL-40 and chitotriosidase inasthma and chronic obstructive pulmonary disease. Am J Respir CritCare Med 193: 131–142. doi:10.1164/rccm.201504-0760oc

Kanneganti M, Kamba A, Mizoguchi E (2012) Role of chitotriosidase (chitinase 1)under normal and disease conditions. J Epithel Biol Pharmacol 5: 1–9.doi:10.2174/1875044301205010001

Kato M, Yuan H, Xu ZG, Lanting L, Li SL, Wang M, Hu MC, Reddy MA, Natarajan R(2006) Role of the Akt/FoxO3a pathway in TGF-beta1-mediatedmesangial cell dysfunction: A novel mechanism related to diabetickidney disease. J Am Soc Nephrol 17: 3325–3335. doi:10.1681/asn.2006070754

Kavsak K, Rasmussen RK, Causing CG, Bonni S, Zhu H, Thomsen GH, Wrana JL(2000) Smad7 binds to Smurf2 to form an E3 ubiquitin ligase thattargets the TGFβ receptor for degradation. Mol Cell 6: 1365–1375.doi:10.1016/s1097-2765(00)00134-9

Kelly M, Kolb M, Bonniaud P, Gauldie J (2003) Re-evaluation of fibrogeniccytokines in lung fibrosis. Curr Pharm Des 9: 39–49. doi:10.2174/1381612033392341

Khalil N, O’Connor RN, Flanders KC, Unruh H (1996) TGF-beta 1, but not TGF-beta 2 or TGF-beta 3, is differentially present in epithelial cells ofadvanced pulmonary fibrosis: An immunohistochemical study. Am JRespir Cell Mol Biol 14: 131–138. doi:10.1165/ajrcmb.14.2.8630262

Khalil N, Parekh TV, O’Connor R, Antman N, Kepron W, Yehaulaeshet T, Xu YD,Gold LI (2001) Regulation of the effects of TGF-beta 1 by activation oflatent TGF-beta 1 and differential expression of TGF-beta receptors(T beta R-I and T beta R-II) in idiopathic pulmonary fibrosis. Thorax 56:907–915. doi:10.1136/thorax.56.12.907

Krein PM, Winston BW (2002) Roles for insulin-like growth factor I andtransforming growth factor-beta in fibrotic lung disease. Chest 122:289S–293S. doi:10.1378/chest.122.6_suppl.289s

Kurt I, Abasli D, Cihan M, Serdar MA, Olgun A, Saruhan E, Erbil MK (2007)Chitotriosidase levels in healthy elderly subjects. Ann N Y Acad Sci1100: 185–188. doi:10.1196/annals.1395.017

Leask A (2006) Scar wars: Is TGFbeta the phantom menace in scleroderma?Arthritis Res Ther 8: 213. doi:10.1186/ar1976

Lee CG, Cho SJ, Kang MJ, Chapoval SP, Lee PJ, Noble PW, Yehualaeshet T, Lu B,Flavell RA, Milbrandt J, et al (2004) Early growth response gene1-mediated apoptosis is essential for transforming growth factorbeta1-induced pulmonary fibrosis. J Exp Med 200: 377–389.doi:10.1084/jem.20040104

Lee CG, Da Silva CA, Dela Cruz CS, Ahangari F, Ma B, Kang MJ, He CH, Takyar S,Elias JA (2011) Role of chitin and chitinase/chitinase-like proteins ininflammation, tissue remodeling, and injury. Annu Rev Physiol 73:479–501. doi:10.1146/annurev-physiol-012110-142250

Lee CG, Hartl D, Lee GR, Koller B, Matsuura H, Da Silva CA, Sohn MH, Cohn L,Homer RJ, Kozhich AA, et al (2009) Role of breast regression protein 39(BRP-39)/chitinase 3-like-1 in Th2 and IL-13-induced tissue responsesand apoptosis. J Exp Med 206: 1149–1166. doi:10.1084/jem.20081271

Lee CG, Herzog EL, Ahangari F, Zhou Y, Gulati M, Peng X, Feghali Bostwick C,Jimenez SA, Varga J, Elias JA (2012) Chitinase 1 is a biomarker for andtherapeutic target in scleroderma-associated interstitial lung diseasethat augments TGF-β1 signaling. J Immunol 189: 2635–2644.doi:10.4049/jimmunol.1201115

Lee CM, He CH, Nour AM, Zhou Y, Ma B, Park JW, Kim KH, Dela Cruz C, Sharma L,Nasr ML, et al (2016) IL-13Ralpha2 uses TMEM219 in chitinase 3-like-1-

induced signalling and effector responses. Nat Commun 7: 12752.doi:10.1038/ncomms12752

Liu G, Friggeri A, Yang Y, Milosevic J, Ding Q, Thannickal VJ, Kaminski N,Abraham E (2010) miR-21 mediates fibrogenic activation of pulmonaryfibroblasts and lung fibrosis. J Exp Med 207: 1589–1597. doi:10.1084/jem.20100035

Massague J (1998) TGF-beta signal transduction. Annu Rev Biochem 67:753–791. doi:10.1146/annurev.biochem.67.1.753

Miyazawa K, Miyazono K (2017) Regulation of TGF-beta family signaling byinhibitory smads. Cold Spring Harb Perspect Biol 9: a022095.doi:10.1101/cshperspect.a022095

Mori T, Kawara S, Shinozaki M, Hayashi N, Kakinuma T, Igarashi A, TakigawaM, Nakanishi T, Takehara K (1999) Role and interaction ofconnective tissue growth factor with transforming growth factor-beta in persistent fibrosis: A mouse fibrosis model. J Cell Physiol181: 153–159. doi:10.1002/(sici)1097-4652(199910)181:1<153::aid-jcp16>3.0.co;2-k

Nakao A, Fujii M, Matsumura R, Kumano K, Saito Y, Miyazono K, Iwamoto I(1999) Transient gene transfer and expression of Smad7 preventsbleomycin-induced lung fibrosis in mice. J Clin Invest 104: 5–11.doi:10.1172/jci6094

Nho RS, Hergert P (2014) FoxO3a and disease progression. World J Biol Chem5: 346–354. doi:10.4331/wjbc.v5.i3.346

Nho RS, Hergert P, Kahm J, Jessurun J, Henke C (2011) Pathological alterationof FoxO3a activity promotes idiopathic pulmonary fibrosis fibroblastproliferation on type I collagen matrix. Am J Pathol 179: 2420–2430.doi:10.1016/j.ajpath.2011.07.020

Nho RS, Peterson M, Hergert P, Henke CA (2013) FoxO3a (Forkhead Box O3a)deficiency protects Idiopathic Pulmonary Fibrosis (IPF) fibroblastsfrom type I polymerized collagen matrix-induced apoptosis viacaveolin-1 (cav-1) and Fas. PLoS One 8: e61017. doi:10.1371/journal.pone.0061017

Peverill W, Powell LW, Skoien R (2014) Evolving concepts in the pathogenesisof NASH: Beyond steatosis and inflammation. Int J Mol Sci 15:8591–8638. doi:10.3390/ijms15058591

Raghu G (1998) Interstitial lung disease: A clinical overview and generalapproach. In Fishman’s Pulmonary Diseases and Disorders. FishmanAP, Elias JA, Fishman JA, Grippi MA, R KL, Senior RM (eds), 1037–1053.New York: Mc-Graw Hill Inc.

Sargent JL, Milano A, Bhattacharyya S, Varga J, Connolly MK, Chang HY,Whitfield ML (2010) A TGFbeta-responsive gene signature isassociated with a subset of diffuse scleroderma with increaseddisease severity. J Invest Dermatol 130: 694–705. doi:10.1038/jid.2009.318

Selman M, King TE, Pardo A (2001) Idiopathic pulmonary fibrosis: Prevailingand evolving hypotheses about its pathogenesis and implications fortherapy. Ann Intern Med 134: 136–151. doi:10.7326/0003-4819-134-2-200101160-00015

Sharma L, Amick AK, Vasudevan S, Lee SW, Marion CR, Liu W, Brady V, Losier A,Bermejo SD, Britto CJ, et al (2018) Regulation and role ofchitotriosidase during lung infection with Klebsiella pneumoniae.J Immunol 201: 615–626. doi:10.4049/jimmunol.1701782

Sime PJ, Xing Z, Graham FL, Csaky KG, Gauldie J (1997) Adenovector-mediatedgene transfer of active transforming growth factor-beta1 inducesprolonged severe fibrosis in rat lung. J Clin Invest 100: 768–776.doi:10.1172/jci119590

Sureshbabu A, Muhsin SA, Choi ME (2016) TGF-beta signaling in the kidney:Profibrotic and protective effects. Am J Physiol Ren Physiol 310:F596–F606. doi:10.1152/ajprenal.00365.2015

Tatler AL, Jenkins G (2012) TGF-beta activation and lung fibrosis. Proc AmThorac Soc 9: 130–136. doi:10.1513/pats.201201-003aw

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 13 of 14

Tzivion G, Dobson M, Ramakrishnan G (2011) FoxO transcription factors;Regulation by AKT and 14-3-3 proteins. Biochim Biophys Acta 1813:1938–1945. doi:10.1016/j.bbamcr.2011.06.002

Van Dyken SJ, Liang HE, Naikawadi RP, Woodruff PG, Wolters PJ, Erle DJ,Locksley RM (2017) Spontaneous chitin accumulation in airways andage-related fibrotic lung disease. Cell 169: 497–509 e413. doi:10.1016/j.cell.2017.03.044

van Eijk M, van Roomen CP, Renkema GH, Bussink AP, Andrews L, BlommaartEF, Sugar A, Verhoeven AJ, Boot RG, Aerts JM (2005) Characterization ofhuman phagocyte-derived chitotriosidase, a component of innateimmunity. Int Immunol 17: 1505–1512. doi:10.1093/intimm/dxh328

Wang C, Gu S, Cao H, Li Z, Xiang Z, Hu K, Han X (2016) miR-877-3p targets Smad7and is associated with myofibroblast differentiation and bleomycin-induced lung fibrosis. Sci Rep 6: 30122. doi:10.1038/srep30122

Webb AE, Kundaje A, Brunet A (2016) Characterization of the direct targets ofFOXO transcription factors throughout evolution. Aging Cell 15:673–685. doi:10.1111/acel.12479

Wurthner JU, Frank DB, Felici A, Green HM, Cao Z, Schneider MD, McNally JG,Lechleider RJ, Roberts AB (2001) Transforming growth factor-betareceptor-associated protein 1 is a Smad4 chaperone. J Biol Chem 276:19495–19502. doi:10.1074/jbc.m006473200

Xu YD, Hua J, Mui A, O’Connor R, Grotendorst G, Khalil N (2003) Release ofbiologically active TGF-(beta)1 by alveolar epithelial cells results inpulmonary fibrosis. Am J Physiol Lung Cell Mol Physiol 285: L527–L539.doi:10.1152/ajplung.00298.2002

Yan X, Liu Z, Chen Y (2009) Regulation of TGF-beta signaling by Smad7. ActaBiochim Biophys Sin (Shanghai) 41: 263–272. doi:10.1093/abbs/gmp018

Yehualaeshet T, O’Connor R, Begleiter A, Murphy-Ullrich JE, Silverstein R,Khalil N (2000) A CD36 synthetic peptide inhibits bleomycin-inducedpulmonary inflammation and connective tissue synthesis in the rat.Am J Respir Cell Mol Biol 23: 204–212. doi:10.1165/ajrcmb.23.2.4089

Zhao J, Crowe DL, Castillo C, Wuenschell C, Chai Y, Warburton D (2000) Smad7is a TGF-beta-inducible attenuator of Smad2/3-mediated inhibitionof embryonic lung morphogenesis. Mech Dev 93: 71–81. doi:10.1016/s0925-4773(00)00281-1

Zhou Y, He CH, Herzog EL, Peng X, Lee CM, Nguyen TH, Gulati M, Gochuico BR,Gahl WA, Slade ML, et al (2015) Chitinase 3-like-1 and its receptors inHermansky-Pudlak syndrome-associated lung disease. J Clin Invest125: 3178–3192. doi:10.1172/jci79792

Zhou Y, Peng H, Sun H, Peng X, Tang C, Gan Y, Chen X, Mathur A, Hu B, SladeMD,et al (2014) Chitinase 3-like 1 suppresses injury and promotesfibroproliferative responses in Mammalian lung fibrosis. Sci TranslMed 6: 240ra276. doi:10.1126/scitranslmed.3007096

License: This article is available under a CreativeCommons License (Attribution 4.0 International, asdescribed at https://creativecommons.org/licenses/by/4.0/).

CHIT1 promotes pulmonary fibrosis Lee et al. https://doi.org/10.26508/lsa.201900350 vol 2 | no 3 | e201900350 14 of 14